Agree for the names.txt, I will include it in the bundle.
About lucene, it's the responsibility of the feature to install lucene bundle. About the version, I can add a version range in the elasticsearch bundle to be sure to use a working lucene version. Another way (it's the one that I will adopt regarding your comment) is the embed the lucene package in the elasticsearch bundle as a private package.
Regards JB On 10/04/2012 02:55 PM, Babak Vahdat wrote:
Hi Here you see the history of that file, however I don't know how important that file is, e.g. would camel-elasticsearch stop working inside Karaf, etc if that file is missing, I don't know. That said to me it's too risky... https://github.com/elasticsearch/elasticsearch/commits/master/src/main/resources/config/names.txt And one question though regarding the missing org/apache/lucene folder (be warned as I'v got ZERO knowledge of the OSGi stuff). How do we know/guarantee that the Apache Lucene version being imported and available at OSGi runtime (e.g. Felix, Equinox, etc.) has the CORRECT/MATCHING version being required by elasticsearch and that it would make elasticsearch work properly out of the box? To be more precise, assume the folks @ elasticsearch have intensively tested and released the version 0.19.9 with the Apache Lucene let's say version X. Then @ OSGi Runtime the Version X+1 is the one which is available and gets loaded into the VM through another bundle being installed & active. However the X+1 has a change in it making it not binary-compatible with the expected version X by elasticsearch 0.19.9 which on it's turn has been intensively tested and verified with Apache Lucene version X embedded inside the JAR! What happens then? Babak -- View this message in context: http://servicemix.396122.n5.nabble.com/VOTE-Apache-ServiceMix-Bundles-2012-10-03-release-tp5714621p5714631.html Sent from the ServiceMix - Dev mailing list archive at Nabble.com.
-- Jean-Baptiste Onofré jbono...@apache.org http://blog.nanthrax.net Talend - http://www.talend.com
